Question explored with the scientific record
Conversation so far
- 1 Which vaccine causes autism?
- 2 What causes autism
Answered on this page ↓
What causes autism
Autism is not one thing with one cause. It is a set of outcomes with many roads in, and the strongest evidence points to genes plus timing, not to any single exposure.
The clearest finding in this evidence is that genes carry most of the risk. Twin and family data put heritability around 83% [12], and a separate review puts it near 90% [13]. But identical twins who share the same DNA can still differ, and whole-genome sequencing in such pairs found real genetic differences between them [1]. So genes are necessary but not sufficient.
The environmental signals that recur are about the mother's immune system during pregnancy, not about vaccines. Maternal infection raises risk about 12%, more when hospitalization is needed [3]. Second-trimester fever raises it, and repeated fever after 12 weeks raises it more [4]. The proposed mechanism is maternal immune activation: cytokines like IL-6 and IL-17 crossing to the fetal brain [6][7][8]. That is a plausible mechanism, not proof of cause in any individual child.
| Exposure | What the record shows |
|---|---|
| Valproate in pregnancy | Autism in about 4.4% of exposed vs 1.5% unexposed [11] |
| Preterm birth | Higher multimorbidity, similar genetic burden [2] |
| Maternal infection | About 12% higher risk [3] |
| Second-trimester fever | Higher risk, more with repeat episodes [4] |
One drug stands out as a real, dose-related cause: sodium valproate, an epilepsy medicine. A Danish cohort found autism in about 4 in 100 exposed children versus 1.5 in 100 unexposed [11], and a systematic review found the same pattern across many studies [10]. Paternal valproate showed no such risk [9]. The MMR question was already settled in the last answer and this evidence does not reopen it [5][14].
My call: autism has no single cause. The best-supported contributors are inherited and spontaneous genetic variation, maternal immune activation in pregnancy, preterm birth, and prenatal valproate. Everything else in this evidence is association, not cause. Confidence: moderate on genes and valproate, low on the rest.
Sources used 14
-
Identifying Genomic Variations in Monozygotic Twins Discordant for Autism Spectrum Disorder Using Whole-Genome Sequencing
This study identifies genomic variations in monozygotic twins discordant for Autism Spectrum Disorder (ASD) using whole-genome sequencing, revealing significant genetic differences and potential ASD risk genes.
DOI: 10.1016/J.OMTN.2018.11.015 -
Prematurity and genetic liability for autism spectrum disorder
The study investigates how prematurity interacts with autism genetic liability using SPARK and SSC cohorts, finding higher multimorbidity in preterm ASD despite similar ASD genetic burden, with birth-era polygenic and de novo analyses and a birth-time machine learning model inte…
DOI: 10.1186/s13073-025-01552-3 -
Maternal infection during pregnancy and risk of autism spectrum disorders: A systematic review and meta-analysis
This systematic review and meta-analysis found that maternal infection during pregnancy is associated with a 12% increased risk of autism spectrum disorder (ASD) in offspring, particularly when hospitalization is required, with variations based on the type of infectious agent an…
DOI: 10.1016/j.bbi.2016.06.005 -
Prenatal fever and autism risk
This study investigates the association between maternal fever during pregnancy and the risk of autism spectrum disorders (ASD) in offspring, finding that second-trimester fever significantly increases ASD risk, particularly with multiple fever episodes after 12 weeks' gestation.
DOI: 10.1038/mp.2017.119 -
Factors associated with Autism Spectrum Disorder: a case-control study in the Lebanese population
In this Lebanese case-control study, bivariate analysis found no significant MMR vaccination-related difference between ASD cases and controls, and the authors concluded that this vaccination was not associated with autism.
DOI: 10.2427/13218 -
Maternal immune activation in neurodevelopmental disorders
This review argues that maternal immune activation, acting through microglial dysregulation and cytokine signaling, raises risk for autism spectrum disorder and schizophrenia.
DOI: 10.1002/dvdy.24612 -
Critical Role of the Maternal Immune System in the Pathogenesis of Autism Spectrum Disorder
A comprehensive literature review synthesizing evidence that maternal immune activation and maternal autoantibodies during pregnancy may contribute to autism spectrum disorder risk in offspring via placental antibody transfer and cytokine-mediated effects (notably IL-6 and IL-17…
DOI: 10.3390/biomedicines8120557 -
Therapeutic Targets for Neurodevelopmental Disorders Emerging from Animal Models with Perinatal Immune Activation
This narrative review argues that astrocyte-derived IL-6, IFITM3 and MMP3, induced by perinatal polyI:C/TLR3 immune activation, are candidate therapeutic targets for schizophrenia and autism spectrum disorder.
DOI: 10.3390/ijms161226092 -
“In the Name of the Father: Risk to the Offspring From Paternal Exposure to Valproate During Conception”
A large Danish nationwide cohort study found no evidence that paternal valproate exposure during spermatogenesis increases the risk of major congenital malformations or neurodevelopmental disorders (including autism) in offspring.
DOI: 10.1177/15357597251317286 -
Risks of neurobehavioral teratogenicity associated with prenatal exposure to valproate monotherapy: a systematic review with regulatory repercussions
A systematic review of prenatal valproate monotherapy exposure showing consistent neurobehavioral and neurocognitive risks in offspring (including lower IQ, ADHD, and autism risk), with dose-related effects and strong regulatory implications to avoid valproate during pregnancy.
DOI: 10.1017/s1092852913000990 -
Valproate use during pregnancy was linked to autism spectrum disorder and childhood autism in offspring
A population-based Danish cohort study found that maternal use of valproate during pregnancy is associated with increased risk of autism spectrum disorder and childhood autism in offspring, while other antiepileptic drugs used as monotherapy did not show the same association.
DOI: 10.7326/0003-4819-159-4-201308200-02013 -
Genetic and environmental factors of schizophrenia and autism spectrum disorder: insights from twin studies
A comprehensive review of twin studies on schizophrenia and autism spectrum disorder that synthesizes genetic heritability, environmental and epigenetic factors, and emerging molecular and cellular models (including iPSC-derived neurons and brain organoids) to understand etiolog…
DOI: 10.1007/s00702-020-02188-w -
“Reports of My Death Were Greatly Exaggerated”: Behavior Genetics in the Postgenomic Era
A comprehensive review of behavioral genetics in the postgenomic era, detailing how genome-wide data, polygenic scores, and exogenous environments reshape our understanding of heritability, gene–environment interplay, and causal inference in psychology.
DOI: 10.1146/annurev-psych-052220-103822 -
Diagnosing autism
This is a narrative clinical review summarizing ASD diagnosis, assessment, and management, citing prior evidence on heritability, recurrence, comorbidity, and interventions rather than presenting original experimental data.
DOI: 10.1016/j.paed.2012.09.008